Keywords
autopsy - maternal vascular malperfusion - fetal vascular malperfusion - heart - weight
sparing - placental pathology
Heart disease is the leading cause of death (COD) in the United States, accounting
for one in five deaths.[1] Early risk stratification is essential for targeting interventions to reduce the
burden of heart disease.[2]
[3] Consistent with the developmental origins of the health and disease framework, the
prenatal period may be a critical window for shaping the long-term risk of heart disease.[4] Human studies show that adverse birth outcomes, including preterm birth, poor fetal
growth, and preeclampsia are associated with long-term offspring cardiovascular morbidity
and mortality.[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13] Further, these adverse birth outcomes have been linked with cardiac remodeling and
dysfunction, as assessed both in early life and adulthood.[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21] These cardiac changes may contribute to increased susceptibility for later-life
cardiovascular disease (CVD).[21] Understanding the mechanisms underlying the association between adverse birth outcomes
and offspring cardiac dysfunction is critical for improving the identification of
those at increased risk of CVD and developing interventions to improve cardiovascular
health across the life course.
While adverse birth outcomes are multifactorial, there is increasing recognition that
abnormal placentation is a significant contributor.[22]
[23] Placental lesions indicative of defective deep placentation are grouped as maternal
vascular malperfusion (MVM).[22]
[24] MVM can be evaluated through placental histologic assessment following delivery
and MVM lesions are identified in up to 50% of those born preterm,[25] 96% of those with poor fetal growth,[26] and 95% of those with early-onset preeclampsia.[27] Other chronic placental pathologies that have been shown to contribute to fetal
growth restriction and preeclampsia, including chronic inflammation and fetal vascular
malperfusion (FVM), might also play a role in cardiac development. FVM is of particular
interest because fetal placental vessels are in direct continuity with the developing
fetal cardiovascular system, making an anatomically plausible link between fetal vascular
disease in the placenta, fetal blood flow alterations, and potential impact on cardiac
development and/or function. Thus, assessment of placental MVM and FVM may yield insights
into underlying mechanisms of altered cardiac development and improve the identification
of those at risk of adverse cardiovascular health across the life course.
Studies investigating MVM in relation to offspring cardiovascular disease risk are
limited by data availability, particularly the availability of placental assessments.
However, one study with routine placental assessment demonstrated that MVM is associated
with elevated blood pressure in childhood among those born early preterm (<34 weeks
gestation).[28] Further, our recent findings demonstrate that stillborn fetuses with a COD attributed
to MVM demonstrated relative sparing of heart weight for gestational age (GA).[29] Building on these findings, the purpose of our analysis was to (1) replicate findings
in a different stillbirth cohort and investigate whether similar patterns are observed
for FVM and (2) to determine whether the severity of MVM is similarly associated with
changes in cardiac structure.
Materials and Methods
Sample Derivation
Data were derived from the Stillbirth Collaborative Research Network (SCRN), a study
of stillbirths and livebirths completed at 59 hospitals across five geographic regions
in the United States from 2005 to 2009.[30] Institutional Review Boards from each center and Data Coordinating and Analysis
Center approved study procedures; for each portion of the study, participants gave
written informed consent. Placental and postmortem examinations were standardized
across institutions, as described previously.[31]
[32] Briefly, placentas were examined fresh, whenever possible, by trained SCRN pathologists,
photographed, and weighed after trimming of the umbilical cord and membranes. Representative
sections for histological analysis included two sections of the umbilical cord, one
section perpendicular to the umbilical cord insertion site, one membrane roll, and
four randomly selected sections of placental parenchyma from a center slice. Fetal
COD was assigned according to Initial Causes of Fetal Death Evaluation, which assigns
stillbirth COD according to a standardized protocol; lesions/pathologies are rated
as 1 (present), 2 (possible COD), or 3 (probable COD).[33] Level of maceration was graded and categorized as grade 0 = no maceration, grade
I = skin desquamation involving ≤1% of the body surface area, grade II = skin desquamation
involving >1 but <5% of the body surface area, grade II = skin desquamation involving
≥5% of the body surface area, grade IV = total brown skin discoloration, and grade
V = mummification. In this analysis, severe maceration was considered either grade
IV or V.
Participants were included in these analyses if they delivered a singleton stillbirth
with a GA of ≥20 weeks and consented to complete placental and postmortem examinations.
Participants were excluded if postmortem stillbirth examination revealed evidence
of structural heart disease, congenital or metabolic cardiac disease, or multiple
or syndromic anomalies.
Analyses of MVM severity in relation to the structural fetal heart measures included
all participants meeting eligibility criteria. For analyses based on COD, participants
were restricted to those with COD attributable to MVM and/or FVM. For comparison,
we included controls with COD due to acute placental infection, as acute processes
close to death are unlikely to impact long-term placental function and/or significantly
change cardiac structure. Participants were assigned to the following groups according
to COD:
-
Compromised maternal circulation: probable (3) COD due to extensive villous (parenchymal)
infarcts;
-
Compromised fetal circulation: probable (3) COD due to compromised fetal microcirculation
(thromboembolism of umbilical vein or villous fetal capillaries and avascular villi,
with evidence of obstruction), umbilical cord entrapment (including nuchal, body,
or shoulder cord with evidence of occlusion and fetal hypoxia), or umbilical cord
stricture (true knots, torsions, or narrowing with thrombi or other obstruction);
and
-
Control: possible (2) or probable (3) COD attributable to placental infection (culture
or PCR proven with placental changes) or chorioamnionitis (culture or PCR proven with
funisitis) with the absence of other probable (3) COD in any category.
Participants with COD outside of these parameters were excluded from analyses comparing
MVM, FVM, and controls.
Weights and Measures
Organ weights and measures were extracted from SCRN autopsy data. GA-adjusted z-scores for each organ were derived from published stillbirth references.[34]
[35]
[36] Relative organ weight/measure to body weight was calculated by subtracting body
weight z-score from organ weight or measure z-score for each fetus.
Lesions of Maternal Vascular Malperfusion
Placental MVM lesions were extracted from SCRN placental pathology data. An established
method for scoring the number of MVM lesions present was used, as shown in [Table 1].[37] The total score for MVM lesions was calculated, and groups were defined as scores
of 0–1 (no/mild MVM), 2–3 (moderate MVM), or ≥4 (severe MVM).
Table 1
Calculation of maternal vascular malperfusion lesion score
|
Freedman–Ernst MVM variables
|
Points
|
SCRN placental examination
|
|
Fibrinoid necrosis/acute atherosis
|
1
|
Fibrinoid necrosis or atherosis/foam cells of membranes or basal plate
|
|
Muscularization of basal plate arterioles
|
1
|
Thick vessel walls (hypertrophy) of decidual vasculature on basal plate
|
|
Basal decidual vascular thrombus
|
1
|
Thrombosis of decidual vasculature on basal plate
|
|
Mural hypertrophy of membrane arterioles
|
1
|
Thick vessel walls (hypertrophic) on free membranes
|
|
Single infarct
|
1
|
Villous parenchymal infarction—focal
|
|
Multiple infarcts, <5% of parenchyma
|
2
|
Villous parenchymal infarction—multiple foci, patchy, or diffuse
|
|
Multiple infarcts, ≥5% of parenchyma
|
2
|
|
Increased syncytial knots
|
1
|
Increased syncytial knots of villous parenchyma—multiple foci or diffuse
|
|
Villous agglutination
|
1
|
Not available
|
|
Increased perivillous fibrin deposition
|
1
|
Fibrinoid deposition of villous parenchyma or intervillous fibrin deposition—multiple
foci or diffuse
|
|
Distal villous hypoplasia/small terminal villous diameters
|
1
|
Distal villous hypoplasia of villous parenchyma—multiple foci or diffuse
|
|
Retroplacental blood/hematoma
|
1
|
Hemorrhage/Clot on basal plate
|
|
Retroplacental blood/hematoma with hemosiderin or infarct
|
2
|
Hemorrhage/Clot on basal plate, with villous infarction
|
|
Small for gestational age (SGA) placenta
|
2
|
Placental weight <10th percentile for GA (isolated SGA placenta not considered MVM)
|
Abbreviation: GA, gestational age; MVM, maternal vascular malperfusion; SCRN, Stillbirth
Collaborative Research Network.
Statistical Analyses
First, demographic and fetal characteristics were compared across groups (compromised
maternal circulation, compromised fetal circulation, control) using Kruskal–Wallis
(continuous) and chi-square (categorical) tests; post hoc comparisons between groups
were completed using Mann–Whitney (continuous) and chi-square (categorical) tests.
Organ z-scores and organ-to-body z-score differences were compared among and between groups using the appropriate tests,
as above.
Organ z-scores and organ-to-body z-score differences were compared among and between MVM lesion categories (0–1, 2–3,
and ≥4) using the methods detailed above. A Mann–Kendall test of trend was conducted
to determine trends in relative z-scores across groups. The relative proportions of MVM lesion categories across our
initial groups (compromised maternal circulation, compromised fetal circulation, and
control) were compared using a chi-square test.
Statistical analyses were completed using RStudio (version 2022.12.0, R version 4.2.2).
A p-value of 0.05 was used to determine statistical significance for primary analyses
testing overall differences across groups. A p-value threshold of 0.0167 (0.05/3) was used to determine statistical significance
for post hoc analyses, which included three comparisons.
Results
Study Sample
Our final sample comprised 329 participants; of these, 76 met the COD criteria for
classification into our three groups (compromised maternal circulation, compromised
fetal circulation, and control). Sample derivation is shown in [Fig. 1]. Among the COD sample, participants classified as controls were more likely to identify
as a racial minority group (57.9%) in comparison to 19% of those with compromised
maternal and 25% of those with compromised fetal circulation ([Table 2]). Those with compromised maternal circulation as a COD had lower birth weights on
average (784 g, standard deviation [SD]: 569) in comparison to those with compromised
fetal circulation (1,720 g, SD: 1,308) and controls (1,125 g, SD: 1,210). Similarly,
28.6% of those with compromised maternal circulation were small for GA (birth weight
<10th percentile for age and sex), though this was not statistically significantly
different from those with compromised fetal circulation (11.1%) or controls (5.3%;
p = 0.12). There were no statistically significant differences in ethnicity, years
of education, or maternal age at delivery.
Fig. 1 This diagram shows the sample selection within the SCRN cohort. MVM, maternal vascular
malperfusion; SCRN, Stillbirth Collaborative Research Network.
Table 2
Demographic characteristics of study sample by cause of death group (n = 76)
|
Measure
|
Control (n = 19)
|
Compromised fetal circulation (n = 36)
|
Compromised maternal circulation (n = 21)
|
p-Value
|
|
Ethnicity (Hispanic)
|
4 (21.1)
|
18 (50.0)
|
8 (38.1)
|
0.11
|
|
Race (minority)
|
11 (57.9)
|
9 (25.0)
|
4 (19.0)
|
0.015
|
|
Education (y)
|
12.6 (2.6)
|
13.9 (2.7)
|
13.1 (2.6)
|
0.3
|
|
Maternal age at delivery (y)
|
27.3 (6.6)
|
27.5 (7.4)
|
28.7 (5.9)
|
0.6
|
|
Fetus sex (female)
|
6 (31.6)
|
18 (50.0)
|
9 (42.9)
|
0.4
|
|
Fetal weight (g)
|
1,125 (1,210)
|
1,720 (1,308)
|
784 (569)
|
0.047
|
|
Small for gestational age infant
|
1 (5.3)
|
4 (11.1)
|
6 (28.6)
|
0.12
|
|
Gestational age at birth (wk)
|
27.0 (7.7)
|
30.9 (6.9)
|
27.7 (4.5)
|
0.056
|
|
Severe maceration (grades IV–V)
|
0 (0)
|
6 (17.1)[a]
|
5 (23.8)
|
0.09
|
Notes: Results are presented as n (%) or mean (standard deviation).
Minority race self-identified as Black, Asian, Native Hawaiian/Pacific Islander, American
Indian/Alaskan Native, or other.
a Maceration grade was missing for one sample.
Among the MVM group, 159 (48.3%) had no/mild MVM, 82 (24.9%) had moderate MVM, and
88 (26.8%) had severe MVM ([Table 3]). The proportion of participants identified as Hispanic differed by MVM severity,
with 28.9% identified as Hispanic among the no/mild MVM group, 34.1% among the moderate
MVM group, and 46.0% among the severe MVM group (p = 0.027). There were no differences in other demographic characteristics, including
race, education, and maternal age. While GA and fetal sex also did not differ across
groups, those with severe MVM had the lowest birth weight (1,006 g, SD: 852, p = 0.022).
Table 3
Demographic characteristics of study sample by the severity of maternal vascular malperfusion,
n = 329
|
Measure
|
Number of MVM lesions
|
p-Value
|
|
0–1 (n = 159)
|
2–3 (n = 82)
|
4+ (n = 88)
|
|
Ethnicity (Hispanic)
|
46 (28.9)
|
28 (34.1)
|
40 (46.0)
|
0.027
|
|
Race (minority)
|
62 (39.0)
|
27 (32.9)
|
27 (30.7)
|
0.4
|
|
Education (y)
|
13.0 (2.7)
|
13.0 (2.9)
|
13.1 (3.0)
|
0.9
|
|
Maternal age at delivery (y)
|
27.6 (6.2)
|
27.0 (6.5)
|
27.9 (6.7)
|
0.8
|
|
Fetus sex (female)
|
72 (45.3)
|
41 (50.0)
|
45 (51.1)
|
0.6
|
|
Fetal weight (g)
|
1,524 (1,237)
|
1,432 (1,350)
|
1,006 (852)
|
0.022
|
|
Gestational age at birth (wk)
|
29.3 (7.0)
|
28.8 (7.1)
|
28.3 (5.7)
|
0.6
|
Abbreviation: MVM, maternal vascular malperfusion.
Notes: Results are presented as n (%) or mean (standard deviation).
Minority race self-identified as Black, Asian, Native Hawaiian/Pacific Islander, American
Indian/Alaskan Native, or Other.
Analysis by Cause of Death
Measures of organ weight z-scores were generally statistically significantly smaller for those with COD related
to compromised maternal circulation ([Table 4]). However, there was no statistically significant difference in brain weight z-score or heart weight z-score across the three groups. Heart measure z-scores were generally smaller than expected for GA in the compromised maternal circulation
group (mean z-scores <0), whereas heart measures were generally as expected or larger than expected
in the compromised fetal circulation and control groups (mean z-scores ≥0). Post hoc
analyses comparing those with compromised maternal circulation to controls demonstrated
statistically significantly smaller left ventricular thickness, tricuspid valve circumference,
mitral valve circumference, and aortic valve circumference.
Table 4
Organ and heart measure z-scores by cause of death (n = 76)
|
Measure
Mean z-score ± SD
|
Cause of death
|
Overall p-value
|
|
Control (n = 19)
|
Compromised fetal circulation (n = 36)
|
Compromised maternal circulation (n = 21)
|
|
Body weight
|
0.11 ± 0.96
|
0.29 ± 1.20
|
−0.86 ± 0.90[b]
[c]
|
<0.001
|
|
Placental weight
|
−0.16 ± 1.90
|
−0.85 ± 3.62
|
−2.57 ± 1.44[b]
[c]
|
<0.001
|
|
Heart weight
|
−0.07 ± 1.10
|
−0.39 ± 0.82
|
−0.71 ± 0.82
|
0.1
|
|
Liver weight
|
0.62 ± 1.15
|
−0.20 ± 1.17[a]
|
−0.94 ± 0.66[b]
[c]
|
<0.001
|
|
Brain weight
|
0.06 ± 0.92
|
−0.10 ± 1.49
|
−0.50 ± 1.10
|
0.5
|
|
Lung weight
|
0.55 ± 1.32
|
0.44 ± 1.31
|
−0.70 ± 0.88[b]
[c]
|
<0.001
|
|
Thymus weight
|
1.62 ± 3.83
|
0.16 ± 1.14
|
−0.26 ± 3.13[b]
[c]
|
<0.001
|
|
Spleen weight
|
0.55 ± 1.37
|
−0.05 ± 0.98
|
−0.81 ± 0.95[b]
[c]
|
<0.001
|
|
Kidney weight
|
0.05 ± 1.12
|
−0.27 ± 1.23
|
−1.11 ± 1.01[b]
[c]
|
0.004
|
|
Adrenal weight
|
0.68 ± 0.97
|
0.51 ± 2.35
|
−0.86 ± 0.75[b]
[c]
|
<0.001
|
|
Right ventricle thickness
|
0.45 ± 1.35
|
0.08 ± 1.06
|
−0.49 ± 0.94
|
0.024
|
|
Left ventricle thickness
|
0.90 ± 1.99
|
−0.26 ± 1.31
|
−0.42 ± 1.23[b]
|
0.011
|
|
Tricuspid valve circumference
|
0.71 ± 1.32
|
0.16 ± 1.50
|
−0.53 ± 1.15[b]
|
0.007
|
|
Pulmonary valve circumference
|
0.31 ± 1.20
|
0.26 ± 1.69
|
−0.58 ± 1.14
|
0.054
|
|
Mitral valve circumference
|
0.64 ± 1.04
|
0.29 ± 1.27
|
−0.11 ± 0.96[b]
|
0.038
|
|
Aortic valve circumference
|
0.44 ± 1.26
|
0.37 ± 2.19
|
−0.70 ± 1.02[b]
|
0.024
|
Abbreviation: SD, standard deviation.
a
p-Value <0.0167 for post hoc comparison of compromised fetal circulation versus control.
b
p-Value <0.0167 for post hoc comparison of compromised maternal circulation versus
control.
c
p-Value <0.0167 for post hoc comparison of compromised fetal circulation versus compromised
maternal circulation.
In analyses accounting for body size (difference between organ weight or measure z-score and body weight z-score), heart weight was as expected for body size in those with compromised maternal
circulation (mean difference in z-score: −0.04, SD: 0.53), which was not significantly
different from the control group (mean: −0.18, SD: 0.88; [Table 5]). In contrast, those with compromised fetal circulation had heart weights that were
small relative to body weight (mean: −0.68, SD: 0.76, p < 0.01). This trend was also observed for brain weight relative to body weight, where
those with compromised maternal circulation and controls had brain weights as expected
for body weight (compromised maternal: 0.07, SD: 1.05; control: −0.03, SD: 0.60),
and those with compromised fetal circulation had small brain weights relative to body
weight (−0.45, SD: 1.47). However, the difference in brain weight z-score was not statistically significant across groups (p = 0.20). Other heart measures relative to body size generally did not differ across
COD groups.
Table 5
Difference between organ or heart measure z-score and body weight z-score by cause of death (n = 76)
|
Measure
Mean difference in z-score ± SD
|
Cause of death
|
Overall p-value
|
|
Control (n = 19)
|
Compromised fetal circulation (n = 36)
|
Compromised maternal circulation (n = 21)
|
|
Placental weight
|
−0.27 ± 1.81
|
−1.14 ± 3.50[a]
|
−1.76 ± 1.51[b]
|
0.009
|
|
Heart weight
|
−0.18 ± 0.88
|
−0.68 ± 0.76
|
−0.04 ± 0.53[c]
|
0.006
|
|
Liver weight
|
0.51 ± 0.94
|
−0.62 ± 0.86[a]
|
−0.18 ± 0.37[b]
[c]
|
<0.001
|
|
Brain weight
|
−0.03 ± 0.59
|
−0.45 ± 1.47
|
0.07 ± 1.05
|
0.2
|
|
Right ventricle thickness
|
0.34 ± 1.54
|
−0.13 ± 1.56
|
0.37 ± 1.22
|
0.3
|
|
Left ventricle thickness
|
0.79 ± 2.21
|
−0.44 ± 1.55
|
0.44 ± 1.60
|
0.038
|
|
Tricuspid valve circumference
|
0.60 ± 1.24
|
−0.13 ± 1.50
|
0.34 ± 1.27
|
0.2
|
|
Pulmonary valve circumference
|
0.20 ± 1.04
|
−0.03 ± 1.41
|
0.28 ± 1.08
|
0.6
|
|
Mitral valve circumference
|
0.56 ± 1.06
|
0.00 ± 1.41
|
0.76 ± 1.19
|
0.2
|
|
Aortic valve circumference
|
0.36 ± 1.14
|
0.08 ± 1.98
|
0.16 ± 0.81
|
0.4
|
Abbreviation: SD, standard deviation.
a
p-Value <0.0167 for post hoc comparison of compromised fetal circulation versus control.
b
p-Value <0.0167 for post hoc comparison of compromised maternal circulation versus
control.
c
p-Value <0.0167 for post hoc comparison of compromised fetal circulation versus compromised
maternal circulation.
Analysis by Maternal Vascular Malperfusion Severity
Body weight and organ-weight z-scores for GA differed by MVM severity in a dose-dependent fashion, with the most
severe MVM (score ≥4) having the smallest organ weights for GA (z-scores generally <0), those with moderate MVM (score 2–3) having organ weights that
were generally slightly smaller than expected or as expected for GA, and those with
no or mild MVM (score 0–1) generally having organ weights that were as expected or
larger than expected for GA ([Fig. 2]). Tests of trend for all organ weights were statistically significant (p < 0.05). Heart measures displayed similar patterns, with the smallest z-scores observed among the severe MVM group and the largest z-scores observed among the no/mild MVM group. Similarly, all tests for trends were
statistically significant.
Fig. 2 Organ and heart measure z-score by severity of maternal vascular malperfusion (MVM), n = 329. Error bars reflect a 95% confidence interval. circ, circumference; LV, left
ventricle; RV, right ventricle.
In analyses comparing organ weights or heart measure z-score relative to body weight
z-score, heart weight did not differ by MVM severity, though the smallest difference
was observed for those with severe MVM (mean difference between heart weight and body
weight z-score: −0.20, SD: 0.95) as compared to those with moderate MVM (−0.78, SD: 3.57)
or no/mild MVM (−0.23, SD: 0.85; [Table 6]). In comparison, brain weight was slightly large relative to body size in those
with the most severe MVM (0.16, SD: 0.84). In general, those with the most severe
MVM had heart measures that were as expected or larger than expected relative to body
weight and those with moderate MVM had heart measures that were generally smaller
than expected relative to body weight.
Table 6
Difference between organ or heart measure z-score and body weight z-score by the severity of maternal vascular malperfusion, n = 329
|
Measure
Mean difference in z-score ± SD
|
Number of MVM lesions
|
Overall p-value
|
|
0–1 (n = 159)
|
2–3 (n = 82)
|
4+ (n = 88)
|
|
Placental weight
|
−0.68 ± 3.48[a]
|
−2.14 ± 3.72[a]
|
−1.93 ± 1.34[b]
|
<0.001
|
|
Heart weight
|
−0.23 ± 0.85
|
−0.78 ± 3.57
|
−0.20 ± 0.95
|
0.6
|
|
Liver weight
|
−0.16 ± 1.38
|
−0.73 ± 3.62
|
−0.11 ± 0.48
|
0.3
|
|
Brain weight
|
−0.15 ± 1.21
|
−0.78 ± 3.95
|
0.16 ± 0.84
|
0.1
|
|
Right ventricle thickness
|
0.05 ± 1.61
|
−0.47 ± 3.74
|
0.22 ± 1.37
|
0.7
|
|
Left ventricle thickness
|
0.00 ± 1.69
|
−0.51 ± 3.87
|
0.23 ± 1.39
|
0.3
|
|
Tricuspid valve circumference
|
−0.13 ± 1.45
|
−0.67 ± 3.63
|
0.01 ± 1.29
|
0.7
|
|
Pulmonary valve circumference
|
−0.28 ± 1.34
|
−1.01 ± 3.62
|
0.00 ± 1.29[c]
|
0.023
|
|
Mitral valve circumference
|
0.05 ± 1.39
|
−0.54 ± 3.83
|
0.33 ± 1.23
|
0.1
|
|
Aortic valve circumference
|
−0.25 ± 1.60
|
−0.76 ± 2.95
|
0.12 ± 1.14[c]
|
0.04
|
Abbreviations: MVM, maternal vascular malperfusion; SD, standard deviation.
a
p-Value <0.0167 for post hoc comparison of 2–3 versus 0–1.
b
p-Value <0.0167 for post hoc comparison of 4+ versus 0–1.
c
p-Value <0.0167 for post hoc comparison of 2–3 versus 4 + .
Discussion
Fetal heart weight sparing may occur in the setting of MVM when evaluated either as
a COD (compromised maternal circulation) or based on severity (≥4 MVM lesions). Stillbirths
with COD attributed to compromised maternal circulation generally had smaller organ
weights and heart measures relative to GA. However, after accounting for fetal body
weight, those with compromised maternal circulation generally had organ weights and
heart measures that were as expected or larger than expected relative to body size.
Similar patterns were observed with MVM severity; those with the most severe MVM generally
had smaller organ weight and heart measures than expected for their GA. Though in
analyses accounting for body size, those in the severe MVM group generally had organ
weights and heart measures that were as expected or larger than expected relative
to body size. These consistent findings are potentially indicative of heart sparing
in MVM. In contrast, heart weight was small relative to GA among those with compromised
fetal circulation.
Our findings are consistent with our previous work in a stillbirth sample, where we
demonstrated that stillbirths with COD attributed to MVM generally had smaller organ
weights and heart measures for GA but also had relative sparing of heart weight and
heart measures when analysis accounted for body size.[29] In prior work, we observed that MVM stillbirths had heart weight z-scores that were 0.3 SDs larger than expected for body weight. In the present analysis,
we found that heart weights were as expected for body weight among stillbirths with
COD attributed to compromised maternal circulation. Similar patterns were observed
for brain weight in the COD analysis, with large brain weights relative to body weight
observed in prior work and brain weights as expected for body weight observed in the
present analysis. These results also build on our previous findings by demonstrating
that with increasing severity of MVM, heart weight/measure sparing was generally increased.
After accounting for body size, we observed that for the most severe cases of MVM,
heart measures were as expected or larger than expected relative to body size.
Findings of generally small organ weights and heart measures among those with severe
MVM are consistent with changes observed in fetal growth restriction, a common consequence
of severe MVM.[24] Brain weight sparing is also a known feature of growth restriction, and our finding
that brain weights are larger relative to body weight in severe MVM as compared to
no or mild MVM is consistent.[38]
[39] Similarly, growth restriction is associated with changes in cardiac structure and
function, including increased wall thickness, hypertrophy, and a more globular shape,
as assessed in infancy and childhood.[14]
[19]
[39]
[40]
[41]
[42]
[43]
[44]
[45] These changes likely contribute to increased CVD risk in adulthood.[10]
[18]
[45]
[46]
[47]
[48] Findings that stillbirths attributed to compromised fetal circulation have smaller
organ measures relative to body size and are also consistent with reported associations
between fetal vascular malperfusion and growth restriction.[49]
[50]
Strengths and Limitations
Strengths and Limitations
Strengths of the analysis include the use of a multisite study with a large sample
of stillbirths. Standardized autopsy and placental pathology methods and reporting
tools were used across study sites. COD was also evaluated consistently across study
sites. Analyses also used measures standardized by GA to account for differences in
the GA distributions across groups. Limitations include the inability to determine
directionality; placental pathology and fetal measures were both assessed following
delivery. Our results do not indicate whether MVM leads to alterations in fetal cardiac
development or whether abnormal cardiac development contributes to the development
of MVM. Our findings in a stillbirth sample also may not be generalizable to live
births with MVM.
Conclusion
Our findings suggest that MVM, both assessed as a COD and based on placental lesion
severity, is associated with alterations in cardiac structure, with as expected or
larger heart measures observed relative to body size in comparison to those with COD
unrelated to MVM or less severe placental MVM burden. Additional research is needed
to understand the underlying mechanisms and establish directionality. Future studies
should also include measures of neonatal cardiac structure and function to investigate
whether findings are consistent in a live birth sample. Consistent findings in a live
birth sample may have important implications for cardiovascular health across the
life course and may help explain observed associations between adverse birth outcomes
and long-term risk of cardiovascular disease.
